sign in sign up
<<
Home , Deuterostome

Molecular Phylogeny and Divergence Times of Deuterostome

Jaime E. Blair and S. Blair Hedges NASA Astrobiology Institute and Department of Biology, The Pennsylvania State University, University Park

The phylogenetic relationships among deuterostome animals have been debated for many , and a diversity of hy- potheses have been proposed based on both morphological and molecular data. Here we have assembled sequences of 217 nuclear-encoded proteins to address specific questions concerning their relationships and times of origin. We recovered significant support for urochordates as the closest relative of with an analysis of 59 proteins (17,400 amino acids) and suggest that the basal position of urochordates found in previous molecular studies may have been the result of long-branch attraction biases. Our results also support , the pairing of with (nine proteins; 2,382 amino acids), and , the pairing of lampreys with hagfish (25 proteins; 6,895 amino acids). In addition, 325 shared proteins (102,110 amino acids) were obtained from the complete genomes of six vertebrates and a urochordate for phylogenetic analysis and divergence time estimation. An evolutionary timescale was estimated using a local (Bayesian) molecular clock method. We found that most major lineages of deuterostomes arose prior to the Explosion of fossils (;520 MYA) and that several lineages had originated before periods of global glaciation in the .

Introduction Deuterostomes are traditionally distinguished from evidence’’ trees combining morphological and molecular other bilaterian animals (i.e., ) by a number data (Zrzavy et al. 1998; Giribet et al. 2000), gene structure of developmental features, including initial planes, (Oda et al. 2002), and multiple protein alignments the fate of the blastopore, and the origin of the (Philippe, Lartillot, and Brinkmann 2005). Interpretations during embryogenesis. Exceptions to these defining charac- of enigmatic fossils, the cornutes and , as stem- teristics have been pointed out and classic affinities ques- and crown-group also led Jefferies to suggest tioned (e.g., Lovtrup 1975; Field et al. 1988). Five major a closer relationship between vertebrates and urochordates groups are usually recognized within Deuterostomia at (Jefferies 1986). Urochordates tend to have long branches the level of : hemichordates, echinoderms, cephalo- in molecular phylogenies, raising the possibility of long- chordates, urochordates, and vertebrates. Other groups such branch attraction as a biasing factor on their phylogenetic as pogonophorans and chaetognaths, once considered deu- position. terostomes (Barnes 1968), have been allied with proto- The remaining deuterostome phyla, hemichordates and stomes based on molecular data (Telford and Holland echinoderms, are considered to be closely related, and this 1993; Wada and Satoh 1994; Winnepenninckx, Backeljau, pairing has been termed Ambulacraria (Metschnikoff 1869). and De Wachter 1995). Despite intensive study, some major Previously, hemichordates were thought to be paraphyletic, aspects of the deuterostome phylogeny remain unresolved, with enteropneusts (acorn worms) more closely related to including the relationships among chordates and between chordates and pterobranchs closer to echinoderms (Barnes and the lobe-finned fish (). In addi- 1968; Nielsen 1995). Differences in larval and adult mor- tion, the quality of the fossil record of deuterostomes varies phologies may have contributed to the confusion as tornaria among phyla. Calcified skeletons in vertebrates and echino- larval characteristics have grouped hemichordates and echi- derms may have increased the fossilization potential of these noderms (e.g., Hyman 1959), while adult pharyngeal clefts two groups compared to hemichordates or cephalochor- and dorsal nerve cords have allied hemichordates with chor- dates. The chitinous tunic of some urochordates has also left dates (e.g., Barnes 1968). More recently, molecular data evidence in the fossil record for the antiquity of this lineage have shown that hemichordates are monophyletic and that (e.g., Chen et al. 2003). pterobranchs may, in fact, be derived from within enterop- Within the deuterostomes, vertebrates, , neusts (Halanych 1995; Cameron, Garey, and Swalla 2000; and urochordates are usually allied as chordates (Chordata) Winchell et al. 2002). A close relationship between hemi- (R. C. Brusca and G. J. Brusca 1990; Nielsen 1995). A close chordates and echinoderms has been supported with total- relationship between vertebrates and cephalochordates, to the evidence trees (Turbeville, Schulz, and Raff 1994; Zrzavy exclusion of urochordates, has traditionally been supported by et al. 1998; Giribet et al. 2000), molecular sequence data morphology (e.g., Maisey 1986). Molecular data have also (Wada and Satoh 1994; Halanych 1995; Castresana et al. recovered a - group (Turbeville, 1998; Bromham and Degnan 1999; Cameron, Garey, and Schulz, and Raff 1994; Wada and Satoh 1994; Cameron, Swalla 2000; Furlong and Holland 2002; Winchell et al. Garey, and Swalla 2000; Winchell et al. 2002), but more 2002), and similarities in Hox gene complements (Peterson recently, some studies have suggested a closer relation- 2004). In addition, a simple marine worm, ship between vertebrates and urochordates based on ‘‘total- bocki, has been allied with hemichordates and echinoderms using molecular data (Bourlat et al. 2003). Within echino- Key words: molecular clock, timetree, genomes, Chordata, Vertebrata. derms, the relationships between the five classes (starfish, sea urchins, brittle stars, sea cucumbers, and sea lilies) E-mail: [email protected]. has not been resolved (Littlewood et al. 1997; Janies Mol. Biol. Evol. 22(11):2275–2284. 2005 doi:10.1093/molbev/msi225 2001), but the consensus view places sea lilies () Advance Access publication July 27, 2005 as the most basal group.

Ó The Author 2005. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please e-mail: [email protected] 2276 Blair and Hedges

While the monophyly of vertebrates has not been Methods questioned, relationships within the phylum have received much attention. Especially controversial has been determin- Protein sequence data were obtained from the public ing the closest relative of tetrapods, which has important database (National Center for Biotechnology Information implications for theories on the conquest of land. Fossil ev- Entrez) for deuterostome and out-group taxa; a idence suggests that an extinct group of lobe-finned fish, the (typically Drosophila) was used as a known out-group to elpistostegalids, is the closest relative of tetrapods (Ahlberg deuterostomes. In the rare case when a protostome sequence and Milner 1994; Long and Gordon 2004). Most molecular was unavailable (,2% of proteins), a more distant out-group, phylogenies have agreed in implicating a lobe-finned fish as either a cnidarian or a fungus, was used. Mitochondrial- the closest relative, but there has been no consensus as to encoded and short (,100 amino acids) proteins were ex- the correct relationship among tetrapods and the two extant cluded. Orthology was assessed for each protein using lineages (lungfish and the coelacanth) based on either mor- reciprocal Blast best hits (Altschul et al. 1997) and by phological (e.g., Rosen et al. 1981; Forey 1986; Northcutt visually inspecting preliminary trees for monophyly of 1986; Schultze 1986) or molecular evidence (Meyer and known groups (e.g., , vertebrates, echinoderms). Wilson 1990; Stock et al. 1991; Meyer and Dolven A total of 217 proteins that individually address some as- 1992; Hedges, Hass, and Maxson 1993; Yokobori et al. pect of deuterostome phylogeny were obtained and ana- 1994; Zardoya and Meyer 1996, 1997; Cao et al. 1998; lyzed during this stage. Individual protein data sets were Zardoya et al. 1998; Tohyama et al. 2000; Venkatesh, aligned (Thompson et al. 1997), and the shape parameter Erdmann, and Brenner 2001; Brinkmann et al. 2004a, of the gamma distribution was estimated to model rate var- 2004b; Takezaki et al. 2004). iation (Yang 1997). Subsets of the total 217 proteins were Among the jawed fish (), cartilaginous then concatenated to address specific relationships and fish () are assumed to be a monophyletic were analyzed using neighbor joining (NJ) (Kumar, Tamura, group closely related to bony fish () (but see and Nei 2004) and maximum likelihood (ML) (Strimmer Martin 2001). An unorthodox scenario placing cartilagi- and von Haeseler 1996) under a Jones Taylor-Thornton nous fish within bony fish (‘‘Pisces’’) has been suggested (JTT) 1 gamma model, and maximum parsimony (MP) (Rasmussen and Arnason 1999a,1999b; Arnason, Gullberg, branch and bound tree searches (Kumar, Tamura, and and Janke 2001; Arnason et al. 2004) but has received Nei 2004). Bootstrapping was used to determine confidence no additional support. Jawless fish (), including values for each node (2,000 replicates for NJ and MP and lampreys and hagfish, are widely recognized as the closest 1,000 puzzling steps for ML). Concatenations did not in- living relatives of gnathostomes (Romer 1966). Morpho- clude any missing data (i.e., all taxa were present for all logical data have previously suggested that lampreys are proteins). Alternative topologies were tested using an more closely related to gnathostomes (Lovtrup 1977; ML test (Yang 1997; Shimodaira and Hasegawa 1999) Hardisty 1979; Maisey 1986), a claim supported by some and P values (P(SH)) were recorded. A list of individual molecular data (Suzuki et al. 1995; Rasmussen, Janke, and proteins used in each concatenation is included in the Sup- Arnason 1998). Most molecular studies, however, have plementary Data (Supplementary Table 1, Supplementary supported a closer relationship between lampreys and hag- Material online), and all protein alignments are available fish (Stock and Whitt 1992; Mallatt and Sullivan 1998; from the author (J.E.B.). Kuraku et al. 1999; Suga et al. 1999; Hedges 2001; Complete genome sequences were obtained for human Mallatt, Sullivan, and Winchell 2001; Delarbre et al. (version 35.1) (Lander et al. 2001; Venter et al. 2001), Mus 2002; Takezaki et al. 2003), which together form the musculus (version 33b.1) (Waterston et al. 2002), Rattus Cyclostomata. norvegicus (version 3d.1) (Gibbs et al. 2004), the domestic In addition to these phylogenetic uncertainties, a central fowl Gallus gallus (version 1.1a) (Hillier et al. 2004), the question in deuterostome evolution concerns the times of pufferfish Takifugu rubripes (version 2c.1) (Aparicio origin of the major lineages and their relationship to the et al. 2002) and Tetraodon nigroviridis (version 1a.1) of fossils. Previous molecular clock (Jaillon et al. 2004), and the intestinalis studies have suggested Precambrian origins for deutero- (version 1.0) (Dehal et al. 2002). Complete genome tran- stomes and other metazoans (Runnegar 1982; Wray, Levinton, scripts were obtained from the Ensembl Genome Browser and Shapiro 1996; Feng, Cho, and Doolittle 1997; Gu 1998; (http://www.ensembl.org/index.html) for all taxa except Wang, Kumar, and Hedges 1999; Nei, Xu, and Glazko Ciona, which was obtained from the Department of Energy 2001; Hedges et al. 2004), although little evidence exists Joint Genome Institute (http://www.jgi.doe.gov). Those in the fossil record prior to the Cambrian Explosion proteins shared among taxa and present in a single copy (;520 MYA). Recent discoveries of Precambrian fossil em- (panorthologs) within each genome were determined (Blair, bryos (Chen et al. 2000) and small bilaterians (Chen et al. Shah, and Hedges 2005). Pairwise orthologs were first iden- 2004) have been dated to ,580 MYA (Condon et al. 2005). tified between genomes and then assembled into three- Here, we have assembled available nuclear-encoded protein member clusters; those clusters containing a single sequence sequence data along with complete genome sequences to ad- per taxon were retained and further intersected with dress phylogenetic relationships among deuterostomes. other three-member clusters until all genomes were com- Molecular divergence times were also estimated using pared. Sequences from the Drosophila melanogaster ge- a Bayesian local clock method, which allows rates to vary nome (version 3b.1) (Adams et al. 2000) were added to among lineages, and robust fossil information to define each cluster to serve as the out-group. A total of 325 shared minimum constraints on each node. proteins were assembled from the intersection of seven Timetree of Deuterostomes 2277 deuterostome genomes; individual data sets were aligned A 81 99 Vertebrates and concatenated as described above. 91 Divergence times were estimated for concatenated data sets using a Bayesian local clock analysis (Kishino, Thorne, Cephalochordate and Bruno 2001) with the following fossil times as mini- mum constraints: mouse-rat 12 MYA (Jacobs and Downs 1994), primate-rodent 65 MYA (Benton 2000), - Urochordate 310 MYA (maximum constraint 370 MYA) (Benton 2000), -hagfish 325 MYA (Donoghue, Smith, and Sansom 2004), - 340 MYA (maxi- mum constraint 370 MYA) (Paton, Smithson, and Clack 0.1 1999; Ruta and Coates 2004), -lobe-finned fish B 100 390 MYA (Benton 1990), tetrapod-ray-finned fish (Actino- 100 Vertebrates ptergyii) 425 MYA (maximum constraint 495 MYA) 88 (Donoghue, Smith, and Sansom 2004), bony fish-cartilaginous fish 457 MYA (Donoghue, Smith, and Sansom 2004), Urochordate starfish-sea urchins 485 MYA (Smith 1988), eleutheroids (starfish 1 sea urchins)-sea lilies 520 MYA (Smith Cephalochordate 1988), agnathans-gnathostomes 520 MYA (Shu et al. 1999), vertebrates-cephalochordates520MYA (Shu, Conway Morris, and Zhang 1996), and vertebrates-urochordates 520 MYA (Chen et al. 2003). As noted, maximum con- 0.1 straints were imposed on three divergences within verte- FIG. 1.—Phylogenetic relationships among phyla. (A) The brates. While the use of maximum constraints may cause traditional relationship among chordates showing the long branch of uro- molecular divergence times to be underestimated (Hedges chordates (59 proteins, a 5 0.58). (B) The influence of using echinoderms et al. 1996; Hedges and Kumar 2004), these constraints as the out-group (19 proteins, a 5 0.56). NJ topologies using a JTT 1 are considered to be conservative based on paleontological gamma model are shown. Bootstrap values for NJ (top), ML (middle), and MP (bottom) are shown above each node. arguments. The presence of Late transitional fos- sils supports the use of 370 MYA as a maximum constraint for the amniote-amphibian divergence (Benton 2000; Clack 2002). Also, the presence of stem-group gnathostomes sup- morphological data (Maisey 1986). Conversely, some re- ports a maximum age of 495 MYA for the origin of ray- cent studies have proposed that urochordates should be finned fish (Donoghue, Forey, and Aldridge 2000; Donoghue, considered the closest relative of vertebrates (Zrzavy et al. Smith, and Sansom 2004). At least one maximum constraint 1998; Giribet et al. 2000; Oda et al. 2002; Philippe, Lartillot, is usually necessary in a Bayesian time estimation analysis; and Brinkmann 2005). With a concatenation of 59 proteins removing all maximum constraints caused divergence times (17,400 amino acids), cephalochordates were significantly to increase by over 50%, a result clearly inconsistent with placed as the closest relative of vertebrates (fig. 1A). How- other aspects of the vertebrate fossil record. ever, the urochordate branch was approximately 50% longer Mean values for the prior distribution of the rate pa- than the vertebrate and cephalochordate branches, making rameter (RtRate in Supplementary Table 1, Supplementary its basal position suspect. The log likelihoods of the two Material online) were estimated for each data set by calcu- possibilities (vertebrates 1 urochordates and vertebrates 1 lating the average in-group branch length (JTT model) and cephalochordates) also were not significantly different dividing by the depth of the in-group node; out-groups were (ÿ178,100.7 vs. ÿ178,121.1 6 26, P(SH) 5 0.23). excluded during timing analyses, and therefore, the use of When an echinoderm was used to root the chordate protostomes as a known out-group to deuterostomes did not phylogeny (19 proteins; 4,877 amino acids), the position affect divergence time estimates. Divergence times were of urochordates changed to become the closest relative of also estimated under a multigene global clock (Hedges vertebrates with strong bootstrap support (fig. 1B), although and Kumar 2003), using a Poisson 1 gamma model and the urochordate branch was approximately twice as long as the mammal-bird split (310 MYA) for rate calibration. the vertebrate and cephalochordate branches. An ML test of this data set significantly rejected the traditional grouping of vertebrates with cephalochordates (P(SH) , 0.01). The re- Results and Discussion covery of this alternative topology when a shorter branched Phylogenetic Analyses and more closely related out-group was used suggests that Subsets of the 217 proteins obtained from the public the phylogenetic position of urochordates may be biased by database were concatenated to address specific phylogenetic long-branch attraction. questions within deuterostomes. Special attention was given To further test for the possibility of long-branch attrac- to the relationships among chordates. Previous molecular tion artifacts, the ratio between the human-urochordate and studies have supported a close relationship between verte- human-cephalochordate pairwise distances was calculated brates and cephalochordates (Turbeville, Schulz, and Raff for each of the 59 proteins used in the initial analysis 1994; Wada and Satoh 1994; Cameron, Garey, and Swalla (fig. 1A). Proteins were then ordered from the highest to 2000; Winchell et al. 2002), a group generally suggested by the lowest ratio values, and those proteins with higher ratios 2278 Blair and Hedges

100 92 100 Mouse

P 80 96 100 60 100 100 Rat 40 100 100 Vertebrate- 20 99

Urochordate Human 97 0 94 88 Domestic Fowl 100 P 47 80 68 Clawed Frog 60 40 85 100 40 98 46 71 Lung Vertebrate- 20 72

Cephalochordate 0 100 Coelacanth 94 1.4 99 Ray-finned Fish 1.2

1.0 Cartilaginous Fish 0.8 100

Average Ratio 0.6 98 Lamprey 0.4 - 59 54 48 42 36 30 24 18 12 6 Number of Proteins

FIG. 2.—The effect on bootstrap support for alternative topologies as 0.05 the ratio between the human-urochordate and human-cephalochordate FIG. 3.—Phylogenetic relationships among vertebrates (14 proteins, pairwise distances decreases. Values on the left side of the graphs represent a 5 0.38). NJ topology using a JTT 1 gamma model is shown. Bootstrap the original concatenation of 59 proteins as presented in fig. 1A. Proteins values for NJ (top), ML (middle), and MP (bottom) are shown above each were then sequentially removed, and bootstrap support (P) for the node. vertebrate-urochordate (top) and vertebrate-cephalochordate (middle) to- pologies was evaluated. The average ratio was calculated for the proteins included in each data set (bottom). also came to similar conclusions (Philippe, Lartillot, and Brinkmann 2005) and suggested that a vertebrate 1 urochor- were sequentially removed in blocks of five to six proteins date relationship should be considered a plausible hypothe- from the concatenation. Bootstrap support for the resulting sis. Further investigation will be needed to fully resolve the NJ topology (Poisson model) was then examined for each relationship among chordates and may be aided by the com- data set (fig. 2). As the ratio between the human-urochordate pletion of the Branchiostoma (cephalochordate) genome. and human-cephalochordate pairwise distances became Phylogenetic analyses of the remaining deuterostome smaller (closer to one), the topology switched from the tra- phyla, echinoderms and hemichordates, supported the ditional grouping of vertebrates and cephalochordates to Ambulacraria hypothesis. With nine proteins, significant a well-supported pairing of vertebrates and urochordates. (.99%) bootstrap support was found with NJ and ML This result suggests that those few proteins with a very high analyses (Supplementary Fig. 2, Supplementary Material ratio (.1.5) had a large influence on the concatenated topol- online). Alternative topologies were also significantly re- ogy. As the relative branch lengths became more similar, jected with an ML test (P(SH) , 0.01). MP did not recover analytical artifacts such as long-branch attraction may have the grouping of echinoderms with hemichordates but instead been overcome to produce the resulting topology. showed weak (68%) support for hemichordates as the sister In addition, a subset of 15 proteins whose ratio was be- group to vertebrates. tween 1.1 and 0.9 was concatenated; the resulting topology Within vertebrates, most generally accepted relation- showed highly significant (.98%) support for urochordates ships were recovered with an analysis of 14 proteins as the closest relative of vertebrates (Supplementary Fig. 1, (2,747 amino acids; fig. 3). A close relationship between Supplementary Material online). The likelihood of this the lungfish and coelacanth, which together were recovered topology was also significantly better than the traditional as the closest relative of tetrapods, received moderate boot- vertebrate 1 cephalochordate hypothesis (P(SH) , 0.02). strap support in this phylogeny. Using a larger data set of 41 These results further suggest that the basal position of uro- proteins (9,475 amino acids) with more limited taxonomic chordates to vertebrates and cephalochordates traditionally sampling, overall bootstrap support for this topology in- found in molecular analyses may be biased by long-branch creased (Supplementary Fig. 3, Supplementary Material attraction. A recent study using multiple protein sequences online), but the log likelihood was not significantly different Timetree of Deuterostomes 2279

a Human Table 1 b Domestic Fowl Bayesian Divergence Time Estimates (MYA) for Clawed Frog Deuterostome Taxa c Number Fossil d of Minimum Divergence Coelacanth a b c e Node Proteins Time Time Ray-finned Fish g a 325 310 (370) 326 (311, 354) Cartilaginous Fish b 191 340 (370) 370d Lamprey h f c 48 390 430 (421, 438) Hagfish d 325 425 (495) 476 (442, 494) i e 65 457 525 (494, 580) Urochordate f 25 325 520 (461, 596) Cephalochordate g 51 520 652 (605, 742) m j h 325 520 794 (685, 918) i 75 520 891 (810, 1,067) k j 19 485 581 (500, 669) l Sea Lily k 3 520 730e l 9 520 876 (725, 1,074) m 71 520 896 (832, 1,022) Global Glaciations Cambrian Explosion a Nodes as presented in figure 4. Proterozoic Phanerozoic b Fossil times used as minimum (maximum in parentheses) constraints are 1000 900 800 700 600 500 400 300 200 100 0 described in Methods. c Million years ago (Ma) Mean of the posterior distribution and the 95% credibility interval are shown for each divergence. d The divergence time between and approached the max- FIG. 4.—A Bayesian timescale for deuterostome evolution. Diver- imum constraint of 370 MYA (divergence time without constraint, 392 [386–398] gence times for each node are presented in table 1. Global glaciations MYA). are hypothesized to have occurred approximately 750–700 MYA (Stur- e The divergence between sea lilies and eleutherozoans (starfish 1 sea urchins tian), 665–635 MYA (Marinoan), and 580 MYA (Varanger/Gaskiers) here) is presented as the midpoint between nodes j and l because of the low number of (Hoffmann et al. 2004). The Cambrian Explosion of fossils occurred ap- proteins available for time estimation. proximately 543–520 MYA. from an alternative hypothesis placing lungfish alone as the with a Bayesian local clock method (fig. 4, table 1). Previous closestrelativeoftetrapods(ÿ67,718.57vs.ÿ67,731.98615, studies have estimated deep, Precambrian divergences be- P(SH) 5 0.32). Not surprisingly, these results were similar tween deuterostomes and protostomes (Runnegar 1982; to those of the study responsible for contributing much of the Wray, Levinton, and Shapiro 1996; Feng, Cho, and Doolittle data (Takezaki et al. 2004); this previous study also sug- 1997; Gu 1998; Wang, Kumar, and Hedges 1999; Nei, Xu, gested that the true relationship between tetrapods and and Glazko 2001; Hedges et al. 2004). Our results suggest the lobe-finned fish may be an unresolvable trichotomy thatdeuterostomesoriginatedpriortoapproximately900MYA and that an excess of multiple hits may be biasing phyloge- and that the first divergence between Ambulacraria and the netic analyses. Apparently, the divergence in question oc- chordates occurred 896 (832–1,022; 95% credibility inter- curred over a very short time period (approximately 40 Myr, val) MYA. Some studies have produced considerably youn- estimatedhere),creatingdifficultyinresolving relationships. ger divergence times among deuterostomes (Aris-Brosou As suggested previously (Takezaki et al. 2004), larger se- and Yang 2002, 2003; Douzery et al. 2004) but suffer from quence data sets (.200 proteins) may be required to resolve methodological biases, causing their results to conflict with this relationship, given the rapid divergences between lobe- other well-supported aspects of the fossil record finned fish and the lineage leading to tetrapods. (Blair and Hedges 2005; Ho et al. 2005). Also within vertebrates, the well-established topology The divergence between echinoderms and hemichor- of tetrapods was significantly supported (88%–100% boot- dates was estimated at 876 (725–1,074) MYA. The large strap support). Cartilaginous fish were recovered as the most credibility interval on this divergence was most likely the basal gnathostomes with moderate bootstrap support (see result of the limited size of the data set (nine proteins; also Supplementary Fig. 4, Supplementary Material online), 2,382 amino acids) and may be more tightly constrained although alternative hypotheses could not be rejected with as additional sequences becomes available. Within the echi- an ML test of 65 proteins (ray-finned fish basal P(SH) 5 0.1, noderms, the divergence between starfish and sea urchins Pisces P(SH) 5 0.52). Cyclostomata, the pairing of lamp- was estimated at 581 (500–669) MYA using 19 proteins reys and hagfish, was significantly supported (see also Sup- (6,813 amino acids); only three sequences were available plementary Fig. 5, Supplementary Material online); an for estimating the divergence time of sea lilies, so this diver- alternative hypothesis suggesting lampreys as the sister gence is presented as the midpoint between the two adjacent group to gnathostomes was significantly rejected with a data nodes (table 1). set of 25 proteins (P(SH) , 0.01). Cephalochordates were estimated to have diverged from the other chordates 891 (810–1,067) MYA with an Divergence Time Analyses analysis of 75 proteins (23,696 amino acids). Using the com- plete genome of C. intestinalis (325 proteins; 102,110 amino The phylogenetic framework of deuterostome evolu- acids), a divergence time of 794 (685–918) MYA was esti- tion presented above was used to estimate divergence times mated for the split between vertebrates and urochordates. 2280 Blair and Hedges

Whilethecredibilityintervals onthesetwodivergencesover- Levinton, and Shapiro 1996; Feng, Cho, and Doolittle 1997; lap, the increase in age between the vertebrate-urochordate Gu 1998; Wang, Kumar, and Hedges 1999; Nei, Xu, and andthevertebrate-cephalochordatesplitswasconsistentwith Glazko 2001; Hedges et al. 2004). The presence of fossil the phylogenetic results presented above. A previous study urochordates (Chen et al. 2003), cephalochordates (Shu, showing a younger age for the vertebrate-cephalochordate Conway Morris, and Zhang 1996), agnathans (Shu et al. split (750 6 32 MYA) did not estimate the vertebrate- 1999), and crinoids (Smith 1988) in Cambrian deposits also urochordate divergence in the analysis (Hedges 2001). argues that the major lineages of deuterostomes had origi- Within vertebrates, the divergence between gnathos- nated prior to the Precambrian-Cambrian boundary. What tomes and agnathans was estimated as 652 (605–742) explains the apparent 300-Myr gap in the deuterostome fos- MYA, which was slightly older than previous estimates be- sil record? Stem lineages could have originated and per- tween 550 and 600 MYA (Wray, Levinton, and Shapiro sisted for long periods of time before the diversification 1996; Kumar and Hedges 1998). However, our result was of modern groups (Hedges 2001). Also, recent geochemical based on 51 proteins (17,160 amino acids), which represents studies have suggested that major shifts in seawater ion con- 5–10 times more data than previous analyses. The diver- centration, particularly an increase in calcium, occurred ap- gence between lampreys and hagfish estimated here at proximately 545 MYA (Brennan, Lowenstein, and Horita 520 (461–596) MYA was similar to a previous estimate 2004). The increased availability of calcium may have of 499 6 37 MYA (Hedges 2001), although different data spurned the evolution of biocalcification in different and methods were used (25 proteins; 6,895 amino acids lineages that had originated much earlier. here). Cartilaginous fish were estimated to have diverged Some groups, such as hemichordates, cephalochor- from bony fish at 525 (494–580) MYA based on 65 proteins dates, and urochordates, had already diverged prior to global (20,537 amino acids), similar to a previous estimate of ‘‘Snowball Earth’’ glaciation events in the . 528 6 56 MYA (Kumar and Hedges 1998). We also esti- As many as three major glacial cycles have been suggested mated that the lineage leading to tetrapods diverged from between approximately 750 and 545 MYA (see Hoffmann the other lobe-finned fish 430 (421–438) MYA, while the et al. 2004 and references therein) (fig. 4), and their prox- coelacanth and lungfish diverged 389 (373–404) MYA. imity to the Cambrian Explosion has led some to suggest However, due to the ambiguity in the phylogenetic analyses, that animal phyla originated as a result of the glaciations this divergence is shown as an unresolved trichotomy (e.g., Hoffman et al. 1998). In contrast, the extent of glaci- (fig. 4). The divergence between amphibians and amniotes, ation into the tropics has been questioned by simulation based on a large data set of 191 proteins (81,117 amino studies (Hyde et al. 2000; Poulsen 2003) and by other ev- acids), approached the maximum constraint of 370 MYA. idence from the geologic record (Condon, Prave, and Benn When this maximum constraint was removed, the diver- 2002). If large areas of open water existed during the Neo- gence time was estimated at 392 (386–398) MYA. proterozoic, long periods of glaciation, even as long as 12 A total of 325 shared proteins (102,110 amino acids) Myr (Bodiselitsch et al. 2005), may not have had such a se- were recovered from the complete genomes of a limited rious impact on global biota. Many eukaryotic lineages cer- number of vertebrates and were used to estimate divergence tainly must have survived major changes in global climate, times (see also Supplementary Fig. 6, Supplementary and fossil evidence has suggested that complex microbiota Material online). Our estimate of the divergence between may have even flourished in some areas during Snowball tetrapods and ray-finned fish at 476 (442–494) MYA was Earth times (Corsetti, Awramik, and Pierce 2003). slightly older than that of a previous study (450 6 36 MYA) (Kumar and Hedges 1998), although the confidence Conclusions intervals of these two estimates overlap. The recent comple- tion of the chicken genome sequence allowed for its use Here we have addressed some specific phylogenetic as a minimum fossil constraint (310 MYA), and the diver- questions within deuterostomes and estimated a molecular gence time between and mammals was estimated at timescale for the divergences among major lineages. We 326 (311–354) MYA. have shown that the traditional (basal) placement of uro- For comparison, multigene global clock analyses were chordates observed in previous molecular studies may be also performed. On average, divergence times among an artifact of long-branch attraction. The finding of a close groups were approximately 20% older under the global relationship between urochordates and vertebrates suggests clock (Supplementary Table 2, Supplementary Material that some interpretations of morphological evolution among online). The use of multiple fossil constraints may explain chordates require reevaluation. Our results also support the why those times estimated with a Bayesian local clock pairing of echinoderms with hemichordates as Ambulacra- methods were younger than those estimated with a single ria, corroborating morphological interpretations of larval (global) evolutionary rate. Divergence time estimates from similarities between these two groups. Within vertebrates, global clock analyses of small (,20 proteins) data sets are lampreys and hagfish were confirmed to be closely related also subject to greater stochasticity, as noted elsewhere (Cyclostomata) and the closest relatives of gnathostomes. (Kumar and Hedges 1998; Hedges et al. 2004). Our analyses suggest that the coelacanth and lungfish form Our Bayesian timescale showed that all major lineages a group that is the closest living relative of tetrapods and that of deuterostomes had originated before the Cambrian Ex- cartilaginous fish are the most basal gnathostomes. Both of plosion in the fossil record (;520 MYA). Previous molec- these results, however, did not receive unambiguous support ular clock studies have also found deep, Precambrian and will require further analysis with larger molecular data divergences between animal phyla (Runnegar 1982; Wray, sets. Our Bayesian timescale suggests that deuterostomes Timetree of Deuterostomes 2281 originated in the Proterozoic and that the major lineages had Bourlat, S. J., C. Nielsen, A. E. Lockyer, D. T. Littlewood, and diverged prior to global glaciation events and the Cambrian M. J. Telford. 2003. Xenoturbella is a deuterostome that eats Explosion of fossils. By establishing well-supported phy- molluscs. Nature 424:925–928. logenies and evolutionary timescales, the interplay between Brennan, S. T., T. K. Lowenstein, and J. Horita. 2004. Seawater geological events and biological innovations can be exam- chemistry and the advent of biocalcification. Geology 32: ined. 473–476. Brinkmann, H., A. Denk, J. Zitzler, J. J. Joss, and A. Meyer. 2004a. Complete mitochondrial genome sequences of the South Amer- Supplementary Material ican and the Australian lungfish: testing of the phylogenetic Supplementary Figures 1–6 and Tables 1 and 2 are performance of mitochondrial data sets for phylogenetic prob- available at Molecular Biology and Evolution online lems in tetrapod relationships. J. Mol. Evol. 59:834–848. Brinkmann, H., B. Venkatesh, S. Brenner, and A. Meyer. 2004b. (http://www.mbe.oxfordjournals.org/). Nuclear protein-coding genes support lungfish and not the coe- lacanth as the closest living relatives of land vertebrates. Proc. Acknowledgments Natl. Acad. Sci. USA 101:4900–4905. We thank David Geiser and two anonymous reviewers Bromham, L. D., and B. M. Degnan. 1999. Hemichordates and deuterostome evolution: robust molecular phylogenetic support for critically reading an earlier version of this manuscript. for a hemichordate 1 echinoderm clade. Evol. Dev. 1:166–171. We also thank Fabia Battistuzzi and Jennifer Sander for as- Brusca, R. C., and G. J. Brusca. 1990. Invertebrates. Sinauer Asso- sistance with data collection. This work was supported by ciates, Sunderland, Mass. grants to S.B.H. from the National Science Foundation and Cameron, C. B., J. R. Garey, and B. J. Swalla. 2000. Evolution of the NASA Astrobiology Institute. the chordate body plan: new insights from phylogenetic anal- yses of deuterostome phyla. Proc. Natl. Acad. Sci. USA 97: Literature Cited 4469–4474. Cao, Y., P. J. Waddell, N. Okada, and M. Hasegawa. 1998. The Adams, M. D., S. E. Celniker, R. A. Holt et al. (195 co-authors). complete mitochondrial DNA sequence of the Mustelus 2000. The genome sequence of Drosophila melanogaster. manazo: evaluating rooting contradictions to living bony ver- Science 287:2185–2195. tebrates. Mol. Biol. Evol. 15:1637–1646. Ahlberg, P. E., and A. R. Milner. 1994. The origin and early Castresana, J., G. Feldmaier-Fuchs, S. Yokobori, N. Satoh, and S. diversification of tetrapods. Nature 368:507–514. Paabo. 1998. The mitochondrial genome of the hemichordate Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, carnosus and the evolution of deuterostome mi- W. Miller, and D. J. Lipman. 1997. Gapped BLAST and tochondria. Genetics 150:1115–1123. PSI-BLAST: a new generation of protein database search Chen, J.-Y., D. J. Bottjer, P. Oliveri, S. Q. Dornbos, F. Gao, S. programs. Nucleic Acids Res. 25:3389–3402. Ruffins, H. Chi, C.-W. Li, and E. H. Davidson. 2004. Small Aparicio, S., J. Chapman, E. Stupka et al. (41 co-authors). 2002. bilaterian fossils from 40 to 55 million years before the Cam- brian. Science 305:218–222. Whole-genome shotgun assembly and analysis of the genome Chen, J.-Y., D.-Y. Huang, Q.-Q. Peng, H.-M. Chi, X.-Q. Wang, 297 of Fugu rubripes. Science :1301–1310. and M. Feng. 2003. The first tunicate from the Early Cambrian Aris-Brosou, S., and Z. Yang. 2002. Effects of models of rate evo- of South China. Proc. Natl. Acad. Sci. USA 100:8314–8318. lution on estimation of divergence dates with special reference Chen, J.-Y., P. Oliveri, C.-W. Li, G.-Q. Zhou, F. Gao, J. W. to the metazoan 18S ribosomal RNA phylogeny. Syst. Biol. Hagadorn, K. J. Peterson, and E. H. Davidson. 2000. Precam- 51:703–714. brian animal diversity: putative phosphatized embryos from ———. 2003. Bayesian models of episodic evolution support the Doushantuo Formation of China. Proc. Natl. Acad. Sci. a Late Precambrian explosive diversification of the Metazoa. USA 97:4457–4462. Mol. Biol. Evol. 20:1947–1954. Clack, J. A. 2002. Gaining ground: the origin and early evolution Arnason, U., A. Gullberg, and A. Janke. 2001. Molecular phylo- of tetrapods. Indiana University Press, Bloomington, Ind. genetics of gnathostomous (jawed) fishes: old , new Condon, D., M. Zhu, S. Bowring, W. Wang, A. Yang, and Y. Jin. cartilage. Zool. Scr. 30:249–255. 2005. U-Pb ages from the Neoproterozoic Doushantuo Forma- Arnason, U., A. Gullberg, A. Janke, J. Joss, and C. Elmerot. 2004. tion, China. Science 308:95–98. Mitogenomic analyses of deep gnathostome divergences: a fish Condon, D. J., A. R. Prave, and D. I. Benn. 2002. Neoproterozoic is a fish. Gene 333:61–70. glacial-rainout intervals: observations and implications. Barnes, R. D. 1968. Invertebrate . 2nd edition. WB Geology 30:35–38. Saunders Co., Philadelphia, Pa. Corsetti, F. A., S. M. Awramik, and D. Pierce. 2003. A complex Benton, M. J. 1990. Phylogeny of the major tetrapod groups: mor- microbiota from Snowball Earth times: microfossils from the Neoproterozoic Kingston Peak Formation, Death Valley, phological data and divergence dates. J. Mol. Evol. 30:409–424. USA. Proc. Natl. Acad. Sci. USA 100:4399–4404. ———. 2000. Vertebrate paleontology. Chapman and Hall, Dehal, P., Y. Satou, R. K. Campbell et al. (87 co-authors). 2002. New York. The draft genome of Ciona intestinalis: insights into chordate Blair, J. E., and S. B. Hedges. 2005. Molecular clocks do not sup- and vertebrate origins. Science 298:2157–2167. port the Cambrian Explosion. Mol. Biol. Evol. 22:387–390. Delarbre, C., C. Gallut, V. Barriel, P. Janvier, and G. Gachelin. Blair, J. E., P. Shah, and S. B. Hedges. 2005. Evolutionary 2002. Complete mitochondrial DNA of the hagfish, Eptatretus sequence analysis of complete eukaryotic genomes. BMC burgeri: the comparative analysis of mitochondrial DNA se- Bioinformatics 6:53. quences strongly supports the cyclostome monophyly. Mol. Bodiselitsch, B., C. Koeberl, S. Master, and W. U. Reimold. 2005. Phylogenet. Evol. 22:184–192. Estimating duration and intensity of Neoproterozoic Snowball Donoghue, P. C., P. L. Forey, and R. J. Aldridge. 2000. Conodont glaciations from Ir anomalies. Science 308:239–242. affinity and chordate phylogeny. Biol. Rev. 75:191–251. 2282 Blair and Hedges

Donoghue, P. C. J., M. P. Smith, and I. J. Sansom. 2004. The Hyde, W. T., T. J. Crowley, S. K. Baum, and W. R. Peltier. 2000. origin and early evolution of chordates: molecular clocks Neoproterozoic ÔSnowball EarthÕ simulations with a coupled and the fossil record. Pp. 190–223 in P. C. J. Donoghue climate/ice-sheet model. Nature 405:425–429. and M. P. Smith, eds. Telling the evolutionary time: molecular Hyman, L. H. 1959. The invertebrates: smaller coelomate groups. clocks and the fossil record. CRC Press, New York. McGraw-Hill, New York. Douzery, E. J. P., E. A. Snell, E. Bapteste, F. Delsuc, and H. Jacobs, L. L., and W. R. Downs. 1994. The evolution of murine Philippe. 2004. The timing of eukaryote evolution: does rodents in Asia. Pp. 149–156 in Y. Tomida, C. K. Li, and a relaxed molecular clock reconcile proteins and fossils? Proc. T. Setoguchi, eds. Rodent and lagomorph families of Asian Natl. Acad. Sci. USA 101:15386–15391. origins and diversification. National Science Museum Feng, D. F., G. Cho, and R. F. Doolittle. 1997. Determining di- Monographs, Tokyo, Japan. vergence times with a protein clock: update and reevaluation. Jaillon, O., J.-M. Aury, F. Brunet et al. (61 co-authors). 2004. Proc. Natl. Acad. Sci. USA 94:13028–13033. Genome duplication in the teleost fish Tetraodon nigroviridis Field, K. G., G. J. Olsen, D. J. Lane, S. J. Giovannoni, M. T. reveals the early vertebrate proto-karyotype. Nature 431: Ghiselin, E. C. Raff, N. R. Pace, and R. A. Raff. 1988. Molecular 946–957. phylogeny of the animal . Science 239:748–753. Janies, D. 2001. Phylogenetic relationships of extant echinoderm Forey, P. 1986. Relationships of lungfishes. J. Morphol. Suppl. classes. Can. J. Zool. 79:1232–1250. 1:75–91. Jefferies, R. P. S. 1986. The ancestry of the vertebrates. British Furlong, R. F., and P. W. H. Holland. 2002. Bayesian phyloge- Museum (Natural History), London. netic analysis supports monophyly of Ambulacraria and of Kishino, H., J. L. Thorne, and W. J. Bruno. 2001. Performance of cyclostomes. Zool. Sci. 19:593–599. a divergence time estimation method under a probabilistic Gibbs, R. A., G. M. Weinstock, M. L. Metzker et al. (228 co- model of rate evolution. Mol. Biol. Evol. 18:352–361. authors). 2004. Genome sequence of the Brown Norway rat Kumar, S., and S. B. Hedges. 1998. A molecular timescale for yields insights into mammalian evolution. Nature 428:493– vertebrate evolution. Nature 392:917–920. 521. Kuraku, S., D. Hoshiyama, K. Katoh, H. Suga, and T. Miyata. Giribet, G., D. L. Distel, M. Polz, W. Sterrer, and W. C. Wheeler. 1999. Monophyly of lampreys and hagfishes supported by 2000. Triploblastic relationships with emphasis on the acoelo- nuclear DNA-coded genes. J. Mol. Evol. 49:729–735. mates and the position of Gnathostomulida, Cycliophora, Pla- Kumar, S., K. Tamura, and M. Nei. 2004. MEGA3: integrated thelminthes, and : a combined approach of 18S software for molecular evolutionary genetics analysis and sequence alignment. Brief. Bioinform. 5:150–163. rDNA sequences and morphology. Syst. Biol. 49:539–562. Lander, E. S., L. M. Linton, B. Birren et al. (255 co-authors). 2001. Gu, X. 1998. Early metazoan divergence was about 830 million Initial sequencing and analysis of the human genome. Nature years ago. J. Mol. Evol. 47:369–371. 409:860–921. Halanych, K. M. 1995. The phylogenetic position of the ptero- Littlewood, D. T., A. B. Smith, K. A. Clough, and R. H. Emson. branch hemichordates based on 18S rDNA sequence data. 1997.Theinterrelationships oftheechinoderm classes: morpho- Mol. Phylogenet. Evol. 4:72–76. logical and molecular evidence. Biol. J. Linn. Soc. 61:409–438. Hardisty, M. W. 1979. Biology of cyclostomes. Chapman and Long, J. A., and M. S. Gordon. 2004. The greatest step in verte- Hall, London. brate history: a paleobiological review of the fish-tetrapod tran- Hedges, S. B. 2001. Molecular evidence for the early history of liv- sition. Physiol. Biochem. Zool. 77:700–719. ing vertebrates. Pp. 119–134 in P. E. Ahlberg, ed. Major events Lovtrup, S. 1975. Validity of the Protostomia-Deuterostomia in early vertebrate evolution. Taylor & Frances Inc., New York. theory. Syst. Zool. 24:96–108. Hedges, S. B., J. E. Blair, M. L. Venturi, and J. L. Shoe. 2004. ———. 1977. The phylogeny of Vertebrata. John Wiley & Sons, A molecular timescale of eukaryote evolution and the rise New York. of complex multicellular life. BMC Evol. Biol. 4:2. Maisey, J. G. 1986. Heads and tails: a chordate phylogeny. Hedges, S. B., C. A. Hass, and L. R. Maxson. 1993. Relations of Cladistics 2:201–256. fish and tetrapods. Nature 363:501–502. Mallatt, J., and J. Sullivan. 1998. 28S and 18S rDNA sequences Hedges, S. B., and S. Kumar. 2003. Genomic clocks and evolu- support the monophyly of lampreys and hagfishes. Mol. Biol. tionary timescales. Trends Genet. 19:200–206. Evol. 15:1706–1718. ———. 2004. Precision of molecular time estimates. Trends Mallatt, J., J. Sullivan, and C. J. Winchell. 2001. The relationship Genet. 20:242–247. of lampreys to hagfish: a spectral analysis of ribosomal DNA Hedges, S. B., P. H. Parker, C. G. Sibley, and S. Kumar. 1996. sequences. Pp. 106–118 in P. E. Ahlberg, ed. Major events in Continental breakup and the ordinal diversification of birds early vertebrate evolution. Taylor & Frances Inc., New York. and mammals. Nature 381:226–229. Martin, A. 2001. The phylogenetic placement of chondrichthyes: Hillier, L. W., W. Miller, E. Birney et al. (175 co-authors). 2004. inferences from analysis of multiple genes and implications for Sequence and comparative analysis of the chicken genome comparative studies. Genetica 111:349–357. provide unique perspectives on vertebrate evolution. Nature Metschnikoff, V. E. 1869. Ueber die systematische stellung von 432:695–716. Balanoglossus. Zool. Anz. 4:139–157. Ho, S. Y., M. J. Phillips, A. J. Drummond, and A. Cooper. 2005. Meyer, A., and S. I. Dolven. 1992. Molecules, fossils, and the Accuracy of rate estimation using relaxed clock models with origin of tetrapods. J. Mol. Evol. 35:102–113. a critical focus on the early metazoan radiation. Mol. Biol. Meyer, A., and A. C. Wilson. 1990. Origin of tetrapods inferred Evol. 22:1355–1363. from their mitochondrial DNA affiliation to lungfish. J. Mol. Hoffman, P. F., A. J. Kaufman, G. P. Halverson, and D. P. Schrag. Evol. 31:359–364. 1998. A Neoproterozoic Snowball Earth. Science 281:1342– Nei, M., P. Xu, and G. Glazko. 2001. Estimation of divergence 1346. times from multiprotein sequences for a few mammalian spe- Hoffmann, K.-H., D. J. Condon, S. A. Bowring, and J. L. Crowley. cies and several distantly related organisms. Proc. Natl. Acad. 2004. U-Pb zircon date from the Neoproterozoic Ghaub For- Sci. USA 98:2497–2502. mation, Namibia: constraints on . Geology Nielsen, C. 1995. Animal evolution: interrelationships of the 32:817–820. living phyla. Oxford University Press, Oxford. Timetree of Deuterostomes 2283

Northcutt, R. G. 1986. Lungfish neural characters and their bearing Suga, H., D. Hoshiyama, S. Kuraku, K. Katoh, K. Kubokawa, and on sarcopterygian phylogeny. J. Morphol. Suppl. 1:277–297. T. Miyata. 1999. Protein tyrosine kinase cDNAs from amphi- Oda, H., H. Wada, K. Tagawa, Y. Akiyama-Oda, N. Satoh, T. oxus, hagfish, and lamprey: isoform duplications around the Humphreys, S. Zhang, and S. Tsukita. 2002. A novel amphi- divergence of cyclostomes and gnathostomes. J. Mol. Evol. oxus cadherin that localizes to epithelial adherens junctions has 49:601–608. an unusual organization with implications for chordate Suzuki, M., K. Kubokawa, H. Nagasawa, and A. Urano. 1995. Se- phylogeny. Evol. Dev. 4:426–434. quence analysis of vasotocin cDNAs of the lamprey, Lampetra Paton, R. L., T. R. Smithson, and J. A. Clack. 1999. An amniote- japonica, and the hagfish, Eptatretus burgeri: evolution of cy- like skeleton from the Early of Scotland. Nature clostome vasotocin precursors. J. Mol. Endocrinol. 14:67–77. 398:508–513. Takezaki, N., F. Figueroa, Z. Zaleska-Rutczynska, and J. Klein. Peterson, K. J. 2004. Isolation of Hox and Parahox genes in 2003. Molecular phylogeny of early vertebrates: monophyly the hemichordate Ptychodera flava and the evolution of of the agnathans as revealed by sequences of 35 genes. deuterostome Hox genes. Mol. Phylogenet. Evol. 31: Mol. Biol. Evol. 20:287–292. 1208–1215. Takezaki, N., F. Figueroa, Z. Zaleska-Rutczynska, N. Takahata, Philippe, H., N. Lartillot, and H. Brinkmann. 2005. Multigene and J. Klein. 2004. The phylogenetic relationship of tetrapod, analyses of bilaterian animals corroborate the monophyly of coelacanth, and lungfish revealed by the sequences of forty- , and Protostomia. Mol. Biol. Evol. four nuclear genes. Mol. Biol. Evol. 21:1512–1524. 22:1246–1253. Telford, M., and P. Holland. 1993. The phylogenetic affinities of Poulsen, C. J. 2003. Absence of a runaway ice-albedo feedback in the chaetognaths: a molecular analysis. Mol. Biol. Evol. 10: the Neoproterozoic. Geology 31:473–476. 660–676. Rasmussen, A. S., and U. Arnason. 1999a. Phylogenetic studies of Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and complete mitochondrial DNA molecules place cartilaginous D. G. Higgins. 1997. The CLUSTALX windows interface: fishes within the tree of bony fishes. J. Mol. Evol. 48:118–123. flexible strategies for multiple sequence alignment aided by ———. 1999b. Molecular studies suggest that cartilaginous fishes quality analysis tools. Nucleic Acids Res. 25:4876–4882. have a terminal position in the piscine tree. Proc. Natl. Acad. Tohyama, Y., T. Ichimiya, H. Kasama-Yoshida, Y. Cao, M. Sci. USA 96:2177–2182. Hasegawa, H. Kojima, Y. Tamai, and T. Kurihara. 2000. Rasmussen, A. S., A. Janke, and U. Arnason. 1998. The mitochon- Phylogenetic relation of lungfish indicated by the amino acid drial DNA molecule of the hagfish (Myxine glutinosa) and sequence of myelin DM20. Mol. Brain Res. 80:256–259. vertebrate phylogeny. J. Mol. Evol. 46:382–388. Turbeville, J., J. Schulz, and R. Raff. 1994. Deuterostome phylog- Romer, A. S. 1966. Vertebrate paleontology. 3rd edition. Univer- eny and the sister group of the chordates: evidence from mol- ecules and morphology. Mol. Biol. Evol. 11:648–655. sity of Chicago Press, Chicago. Venkatesh, B., M. V. Erdmann, and S. Brenner. 2001. Molecular Rosen, D. E., P. L. Forey, B. G. Gardiner, and C. Patterson. 1981. synapomorphies resolve evolutionary relationships of extant Lungfishes, tetrapods, paleontology, and plesiomorphy. Bull. jawedvertebrates. Proc.Natl.Acad.Sci.USA98:11382–11387. Am. Mus. Nat. Hist. 167:159–276. Venter, J. C., M. D. Adams, E. W. Myers et al. (274 co-authors). Runnegar, B. N. 1982. A molecular-clock date for the origin of 2001. The sequence of the human genome. Science 291:1304– animal phyla. Lethaia 15:199–205. 1351. Ruta, M., and M. I. Coates. 2004. Bones, molecules, and crown- Wada, H., and N. Satoh. 1994. Details of the evolutionary tetrapod origins. Pp. 224–262 in P. C. J. Donoghue and M. P. history from invertebrates to vertebrates, as deduced from Smith, eds. Telling the evolutionary time: molecular clocks and the sequences of 18S rDNA. Proc. Natl. Acad. Sci. USA the fossil record. CRC Press, New York. 91:1801–1804. Schultze, H. P. 1986. Dipnoans as sarcopterygians. J. Morphol. Wang, D. Y., S. Kumar, and S. B. Hedges. 1999. Divergence time Suppl. 1:39–74. estimates for the early history of animal phyla and the origin of Shimodaira, H., and M. Hasegawa. 1999. Multiple comparisons of plants, animals and fungi. Proc. R. Soc. Lond. B Biol. Sci. log-likelihoods with applications to phylogenetic inference. 266:163–171. Mol. Biol. Evol. 16:1114–1116. Waterston, R. H., K. Lindblad-Toh, E. Birney et al. (222 co- Shu, D.-G., S. Conway Morris, and X. L. Zhang. 1996. A - authors). 2002. Initial sequencing and comparative analysis like chordate from the Lower Cambrian of China. Nature of the mouse genome. Nature 420:520–562. 384:157–158. Winchell, C. J., J. Sullivan, C. B. Cameron, B. J. Swalla, and J. Shu, D.-G., H.-L. Luo, S. Conway Morrise, X.-L. Zhang, S.-X. Mallatt. 2002. Evaluating hypotheses of deuterostome phylog- Hu, L. Chen, J. Han, M. Zhu, Y. Li, and L.-Z. Chen. 1999. eny and chordate evolution with new LSU and SSU ribosomal Lower Cambrian vertebrates from south China. Nature DNA data. Mol. Biol. Evol. 19:762–776. 402:42–46. Winnepenninckx, B., T. Backeljau, and R. De Wachter. 1995. Smith, A. B. 1988. Fossil evidence for the relationships of extant Phylogeny of protostome worms derived from 18S rRNA echinoderm classes and their times of divergence. Pp. 85–97 in sequences. Mol. Biol. Evol. 12:641–649. C. R. C. Paul, and A. B. Smith, eds. Echinoderm phylogeny Wray, G. A., J. S. Levinton, and L. H. Shapiro. 1996. Molecular and evolutionary biology. Clarendon Press, Oxford. evidence for deep Precambrian divergences among metazoan Stock, D. W., K. D. Moberg, L. R. Maxson, and G. S. Whitt. 1991. phyla. Science 274:568–573. A phylogenetic analysis of the 18S ribosomal RNA sequence Yang, Z. 1997. PAML: a program package for phylogenetic anal- of the coelacanth Latimeria chalumnae. Environ. Biol. ysis by maximum likelihood. Bioinformatics 13:555–556. 32:99–117. Yokobori, S., M. Hasegawa, T. Ueda, N. Okada, K. Nishikawa, Stock, D. W., and G. S. Whitt. 1992. Evidence from 18S ribo- and K. Watanabe. 1994. Relationship among coelacanths, somal RNA sequences that lampreys and hagfishes form a nat- lungfishes, and tetrapods: a phylogenetic analysis based on ural group. Science 257:787–789. mitochondrial cytochrome oxidase I gene sequences. J. Mol. Strimmer, K., and A. von Haeseler. 1996. Quartet puzzling: a quar- Evol. 38:602–609. tet maximum-likelihood method for reconstructing tree topol- Zardoya, R., Y. Cao, M. Hasegawa, and A. Meyer. 1998. Search- ogies. Mol. Biol. Evol. 13:964–969. ing for the closest living relative(s) of tetrapods through 2284 Blair and Hedges

evolutionary analyses of mitochondrial and nuclear data. Mol. Zrzavy, J., S. Mihulka, P. Kepka, A. Bezdek, and D. Tietz. 1998. Biol. Evol. 15:506–517. Phylogeny of the metazoa based on morphology and 18S Zardoya, R., and A. Meyer. 1996. Evolutionary relationships of the ribosomal DNA evidence. Cladistics 14:249–285. coelacanth, lungfishes, and tetrapods based on the 28S ribo- somal RNA gene. Proc. Natl. Acad. Sci. USA 93:5449–5454. Takashi Gojobori, Associate Editor ———. 1997. The complete DNA sequence of the mitochondrial genome of a ‘‘living fossil,’’ the coelacanth (Latimeria chalum- nae). Genetics 146:995–1010. Accepted July 22, 2005